HIPPOCAMPUS 24:869–876 (2014)
Ion Channels of the Nuclear Membrane of Hippocampal Neurons Olena A. Fedorenko,1,2 and Sergey M. Marchenko1,2*
ABSTRACT: Rise in Ca21 concentration in the nucleus affects gene transcription and has been implicated in neuroprotection, transcriptiondependent neuronal plasticity, and pain modulation, but the mechanism of regulation of nuclear Ca21 remains poorly understood. The nuclear envelope is a part of the endoplasmic reticulum and may be one of the sources of nuclear Ca21. Here, we studied ion channels in the nuclear membrane of hippocampal neurons using the patch-clamp technique. We have found that the nuclear membrane of CA1 pyramidal and dentate gyrus granule (DG), but not CA3 pyramidal neurons, was enriched in functional inositol 1,4,5-trisphosphate receptors/Ca21-release channels (IP3Rs) localized mainly in the inner nuclear membrane. A single nuclear ryanodine receptor (RyR) has been detected only in DG granule neurons. Nuclei of the hippocampal neurons also expressed a variety of spontaneously active cation and anion channels specific for each type of neuron. In particular, large-conductance ion channels selective for monovalent cations (LCC) were coexpressed with IP3Rs. These data suggest that: (1) the nuclear membranes of hippocampal neurons contain distinct sets of ion channels, which are specific for each type of neuron; (2) IP3Rs, but not RyRs are targeted to the inner nuclear membrane of CA1 pyramidal and DG granule, but they were not found in the nuclear membranes of CA3 pyramidal neurons; (3) the nuclear envelope of these neurons is specialized to release Ca21 into the nucleoplasm which may amplify Ca21 signals entering the nucleus from the cytoplasm or generate Ca21 transients on its own; (4) LCC channels are an integral part the of Ca21-releasing machinery providing a route for counterflow of R1 and thereby facilitating Ca21 movement in and out of the Ca21 C 2014 Wiley Periodicals, Inc. store. V KEY WORDS: ion channels; cell nucleus; endoplasmic reticulum; calcium stores; hippocampus
INTRODUCTION The nuclear envelope is a specific organelle, which separates the genetic apparatus of a cell from the rest of the cytoplasm. It is a part of the endoplasmic reticulum (ER), connected to it by its outer nuclear
membrane and common luminal space. The relatively large size of the nuclear envelope enables direct patchclamp recording from the nuclear membranes. This approach has been successfully employed for investigation of inositol 1,4,5-trisphosphate receptors (IP3Rs) (Mak et al., 1998; Marchenko et al., 2003; Rahman et al., 2009). A variety of other ion channels was reported in the nuclear membrane (Mazzanti et al., 2001; Fedorenko et al., 2010; Fedorenko and Marchenko, 2010). Besides being a useful model of the ER, the nuclear envelope is involved in the regulation of nuclear Ca21. A rise in nuclear Ca21 concentration ([Ca21]n) affects the transcription of a large pool of genes in hippocampal neurons and has been implicated in various physiological phenomena including neuronal survival, aging and transcription-dependent long-lasting changes in synaptic efficiency (Papadia et al., 2005; Zhang et al., 2007, 2009; Dick and Bading, 2010; Mauceri et al., 2011; Bengtson and Bading, 2012; Oliveira et al., 2012). We have previously reported multiple functional IP3Rs and spontaneously active cation and anion channels in the nuclear membranes of cerebellar Purkinje and granule neurons (Marchenko et al., 2005). Expression of ion channels greatly varied in different types of cerebellar neurons and we, for example, have not detected IP3-activated channels in the nuclear membrane of cerebellar granule neurons. Therefore, every cell type has to be studied independently. Here, we studied ion channels in the membranes of nuclei isolated from CA1 and CA3 pyramidal and dentate gyrus (DG) granule neurons of the rat hippocampus.
MATERIALS AND METHODS 1
Bogomoletz Institute of Physiology, Department of Brain Physiology, 4 Bogomoletz Street, Kiev, 01024, Ukraine; 2 State Key Laboratory of Molecular and Cellular Biology, 4 Bogomoletz Street, Kiev, 01024, Ukraine Additional Supporting Information may be found in the online version of this article. Grant sponsor: The State Fund for Fundamental Researches; Grant number: DFFD F 46.2/001. *Correspondence to: Sergey M. Marchenko, Bogomoletz Institute of Physiology, Department of Brain Physiology, 4 Bogomoletz Street, Kiev 01024, Ukraine. E-mail: [email protected] Accepted for publication 31 March 2014. DOI 10.1002/hipo.22276 Published online 25 April 2014 in Wiley Online Library (wileyonlinelibrary.com). C 2014 WILEY PERIODICALS, INC. V
Isolation of Neuronal Nuclei All experimental procedures involving animals and their care were conducted in accordance with the European Communities Council Directive of 24 November 1986. Hippocampi were isolated from three-week-old male Wistar rats. Transverse slices of the hippocampus were cut by hand and each slice was placed in an eppendorf tube with 1 ml of ice-cold solution containing (mM): potassium gluconate 150, Hepes-KOH 10 (pH 7.2). Protease Inhibitor Cocktail (Roche Diagnostics Gmbh) was added to the solution
870
FEDORENKO AND MARCHENKO
according to the manufacturer’s instructions. The eppendorf tubes with hippocampal slices were placed in a freezer and used for experiments during the next 1–3 days. The pyramidal layers of the CA1 or CA3 areas or DG granule layer were carefully cut using microsurgical instruments. The resulting preparation was homogenized by passing through a 23 gauge needle. The homogenate was centrifuged, the pellet was resuspended in the same solution and placed into a working chamber on the stage of a Leica DMIRB inverted microscope (Leica Microsystems, Germany). The nuclei were allowed to attach to the glass bottom of the chamber and then debris was removed by washing with solution (hereafter referred to as KCl solution) containing (mM): KCl 150; HEPES 5, pH 7.2. Nuclei in the working chamber were visualized with the VX44 or Pixelfly CCD camera (PCO, Kelheim, Germany). To access the inner nuclear membrane the isolated nuclei were treated with 1% (w/v) sodium citrate (Humbert et al., 1996) and incubated for 15–20 min on ice while gently stirring (HI 200M magnetic stirrer, Hanna Instruments, 0.5–1 revolutions per second). They were then used in experiments as described above.
Electrophysiological Studies Single ion channels were recorded from nucleus-attached and excised patches of the nuclear membrane in the voltageclamp mode of the patch-clamp technique. There were no differences between data obtained in these configurations, so the data were pooled. Patch pipettes were prepared from borosilicate glass (Sutter Instrument, Novato, CA, USA) and filled, unless stated otherwise, with KCl solution. Pipette resistance ranged from 8 to 12 MX. Data acquisition was carried out using a Visual Patch VP-500 amplifier (Bio-Logic, Claix, France). Currents were filtered with a low-pass Bessel filter at 2 kHz, digitized at 10 kHz, and stored on a computer disk. All experiments were carried out at room temperature (18–20 C). In all recordings in this paper, the potential of the bath solution was considered to be 0 mV. Data are presented as the mean 6 standard error (s.e.m.) and n (number of observations). To determine the selectivities of ion channels, KCl solution was replaced with either K-gluconate solution (potassium gluconate 150 mM, HEPES 5, pH 7.2) or N-methyl-D-glucamine (NMDG) solution (NMDG-Cl 150 mM, HEPES 5, pH 7.2). Ca21 concentration was buffered with EGTA: EGTA 1 mM, CaCl2 0.6 mM; calculated free Ca21 concentration was 0.26 m. Solutions with agonists of IP3Rs or RyRs also contained ATP (0.5 mM).
RESULTS Isolated Nuclei of Hippocampal Neurons The heterogenous cell composition of the brain complicates isolation of nuclei from specific types of neurons. The regular anatomical structure of the hippocampus helps to overcome Hippocampus
FIGURE 1. Isolated nuclei of hippocampal neurons and anion channels in the outer nuclear membrane of CA1 pyramidal neurons. Cell nuclei isolated from CA1 (A) and CA3 (B) pyramidal neurons of the hippocampus and from granule cells of the dentate gyrus (C). CCD images of unstained nuclei in KCl solution, objective 3100. (D) Recording of the anion channels of the outer nuclear membrane of CA1 pyramidal neurons (an arrow indicates the close state of the channel). (E) A current–voltage relationship of anion channels in the symmetric 150 mM KCl solution (closed circles), and with low Cl– solution (130 mM K-Glu and 20 mM KCl) in the bath solution (open circles).
this problem. The CA1 and CA3 pyramidal layers of Ammon’s Horn and the granule layer of the Dentate Gyrus (DG) are composed of densely packed cell bodies of pyramidal and granule neurons respectively, with few other cell types (Amaral and Lavenex, 2007). In accordance with this, nuclei isolated from CA1, CA3 pyramidal, or DG layers were morphologically homogenous. Nuclei from CA1 pyramidal neurons were spherical in shape and 10–12 mm in diameter. They had one or more distinct dark nucleoli and clear light nucleoplasm (Fig. 1A). These morphological features and size set them apart from the nuclei of other cell types in the CA1 area. The nuclei isolated from CA3 pyramidal neurons were larger (15 lm), had spherical or ellipsoidal shape, several (2–4) clearly visible dark nucleoli and less light nucleoplasm (Fig. 1B). Nuclei from DG granule neurons had diameters of 7–8 mm, small nucleoli, and turbid nucleoplasm (Fig. 1C).
NUCLEAR ION CHANNELS OF HIPPOCAMPAL NEURONS
871
Patch-clamp recordings of the nuclear membranes from these three types of hippocampal neurons revealed multiple ion channels, that varied in their selectivity, conductance, and pattern of activity.
Ion Channels in the Nuclear Membranes of CA1 Pyramidal Neurons In the outer nuclear membrane of ff1 pyramidal neurons, we recorded spontaneously active ion channels with conductances of 156 6 4 pS (n 5 3) in symmetric Kl solution (Fig. 1E). The density of these channels was low (3 out of 56 patches). The channels demonstrated clear voltage-dependence (Fig. 1D). At positive potentials (and therefore negative potentials at the luminal side of the membrane) the open probability of these channels was low (0.02 at 140 mV), but significantly increased at negative potentials (0.08 at 240 mV). Replacement of 80% (130 mM) of KCl in the bath solution with potassium gluconate shifted the reversal potential of the current to 49.3 mV (Fig. 1E). The calculated reversal potential for Cl2 ions in these conditions was 50.9 mV. Replacement of KCl solution in the working chamber with NMDG-Cl solution did not affect the channel. Therefore, the channels were permeable for Cl2 and practically impermeable to K1. No such channels were detected in the inner nuclear membrane. The most common spontaneously active ion channels found in the both inner and outer nuclear membranes of CA1 pyramidal neurons were large-conductance cation (LCC) channels. These channels were mainly localized in the inner nuclear membrane. Out of 151 patches of the inner nuclear membrane, 56 contained one or more (up to 7, mean 3.8 channels per patch) of these channels (37% of patches) and very few were found in the outer nuclear membrane (1 out of 56, or less than 2%). The channels demonstrated clear voltagedependence. At positive potentials at the luminal side of the membrane, they were mainly in the open state and their open probability (Po) was 0.8. At negative potentials, the activity of the channels significantly declined (Fig. 2A). Long exposure to negative potentials below 240 mV could completely, but reversibly block the channels. In symmetric KCl solution, the slope conductance of the channel was 248 6 6 pS (n 5 8) (Fig. 2B). To determine ion selectivity of the channel, KCl solution in the working chamber was replaced with NMDG solution; the patch pipettes contained the standard KCl solution. Under these conditions, only outward currents were recorded over the range of potentials from 280 mV to 180 mV. Equimolar replacement of KCl in the outer solution with NaCl did not noticeably affect the channel. When patch pipettes were filled with 100 mM BaCl2 or CaCl2 solution, only inward currents were observed. These data suggest that LCC channels were selective to monovalent (K1 and Na1) cations and impermeable to divalent (Ca21, Ba21) cations. Agonists of IP3Rs (IP3 1–10 mM, Ca21 0.26 mM, and ATP 0.5 mM) applied to the nucleoplasmic (bath) side of the inner nuclear membrane of CA1 pyramidal neurons activated highconductance channels in 25% of patches (37 out of 151
FIGURE 2. Large conductance cation (LCC) channels in the inner nuclear membrane of CA1 pyramidal neurons. (A) LCC channel recorded at holding potentials of 160 mV and 260 mV. (B) A current–voltage relationship of LCC channels in symmetric KCl solution (closed circles), with NMDG solution in the bath (triangles), and with 100 mM CaCl2 in the patch-pipette (open circles). a21 either in pipette or bath solution at large concentrations (10 mM) reduced the channel amplitude.
patches, on the average 2.1 channels per patch). In the outer nuclear membrane of CA1 neurons, agonists of IP3Rs in the pipette solution (the cytoplasmic side of the outer nuclear membrane) activated channels of the same conductance only in a few patches (3 out of 54 patches, or 5.6%). These channels often overlapped with LCC channels in the same patch. IP3 activated the channels at concentrations 50 nM and saturated their activity (Po 95% of Po at 10 mM IP3) at concentrations of 2 mM (Fig. 3A). When added to the patch pipette solution (the luminal side of the inner nuclear membrane), the agonists did not activate any channels (n 5 37). So, the ligandbinding domains of these IP3Rs were directed toward the nucleoplasm. In symmetric KCl solution, IP3/Ca21-activated channels had linear current–voltage relationships with slope conductances of 366 6 5 pS (n 5 6, Fig. 3B, closed circles). When patch pipettes were filled with solution containing 50 mM CaCl2 and 30 mM KCl, the reversal potential of IP3-activated channels was 224.3 6 2.1 mV (n 5 4, Fig. 3B, open circles). This reversal potential corresponds to a relative permeability of the channel to Ca21 (PCa/PK) of 6.6 (Miedema, 2002). IP3-activated channels were also permeable to Ba21. When patch pipettes were filled with solution containing 100 mM BaCl2, the reversal potential of IP3-activated channels was Hippocampus
872
FEDORENKO AND MARCHENKO nuclear membrane. Large-conductance spontaneously active ion channels, resembling LCC channels of CA1 pyramidal neurons, were recorded in two patches of the outer membrane and two patches of the inner nuclear membrane. In symmetric KCl solution, the channel conductance was 210 6 6 pS (n 5 4). The channels were voltage-dependent, being activated by positive potentials and reversibly inhibited by negative potentials at the luminal side of the nuclear membrane. The rarity of these channels did not allow us to study their properties in more detail. To record Ca21 channels, the pipettes or bath were filled with solutions that contained agonists of RyRs (ryanodine 10 nM) or IP3Rs (IP3, 2 mM; Ca21, 0.26 mM). Two IP3Rs were recorded in a single patch (out of 75) of the outer nuclear membrane. No IP3R or RyRs were detected in the inner nuclear membrane (n 5 86).
Ion Channels in the Nuclear Membranes of DG Granule Neurons
FIGURE 3. IP3Rs in the inner nuclear membrane of CA1 pyramidal neurons. (A) IP3Rs were activated by IP3 in the presence of Ca21 (250 nM), the patch contained at least two IP3Rs, holding potential was 140 mV. (B) A current–voltage relationship of IP3Rs of CA1 pyramidal neurons in symmetric KCl solution was linear with a slope conductance of 366 6 5 pS (closed circles); when patch pipettes were filled with high Ca1-containing solution (50 mM CaCl2 and 30 mM KCl) the reversal potential of IP3/ Ca21-activated channels was 224.3 6 2.1 mV indicating Ca21 selectivity of the channel with PCa/PK 5 6.6 (open circles).
233.1 6 2.8 mV (n 5 4, not illustrated). The calculated PBa/ PK was 6.5. Therefore, IP3/Ca21-activated nuclear channels in CA1 pyramidal neurons can be identified as IP3Rs with properties similar to those previously reported for IP3Rs (Bezprozvanny, 2005; Marchenko et al., 2005). Ryanodine receptors/Ca21-release channels were reported to be expressed in CA1 pyramidal neurons (Sharp et al., 1993, 1999; Hertle and Yeckel, 2007). Nevertheless the agonists of RyRs, cADP ribose (10 mM) or ryanodine (10 nM), added to both pipette and bath solutions did not activate any channels in either the inner or the outer nuclear membrane of CA1 pyramidal neurons (n > 40 for each membrane).
Ion Channels in the Nuclear Membranes of CA3 Pyramidal Neurons The nuclear membranes of CA3 pyramidal neurons were characterized by extremely low densities of ion channels. We detected channel activity in only 3 out of 198 patches of the outer membrane and in only 2 out of 204 patches of the inner Hippocampus
Patch-clamp recording from the nuclear membranes of DG granule neurons revealed a variety of ion channels with densities lower than in the CA1 area, but higher than in CA3. A channel with high amplitude was recorded in the outer nuclear membrane of granule neurons (Fig. 4A). In symmetric KCl solution, it had a slope conductance of 305 6 16 pS (n 5 5, Fig. 4B, closed circles). Substitution of KCl solution with NMDG-Cl solution did not affect the channel, but with R1gluconate in the bath only inward currents were observed (Fig. 4B, open circles), suggesting that the channel was selective for Cl2. The most common spontaneously active channels in the nuclear membranes of DG granule neurons were largeconductance ion channels. These channels by their pattern of activity, high amplitude, slow kinetics, voltage-dependence, and localization mainly in the inner nuclear membrane resembled LCC channels of CA1 pyramidal neurons, but had smaller conductance. In symmetric KCl solution, the slope conductance of the channels was 179 6 15 pS (n 5 6). Equimolar substitution of KCl in the bath solution with K-gluconate did not affect the channels, but when KCl bath solution was replaced with NMDG-Cl solution only outward currents were observed over the range of potentials from 280 mV to 180 mV. With CaCl2 (100 mM) in the pipette only outward currents were detected. Therefore, the channel was permeable to K1, but impermeable to Ca21, Cl2, and NMDG. One more spontaneously active ion channel was recorded in the inner nuclear membrane of granule neurons. The channel quickly fluctuated between two-subconductance states (Fig. 5A). In symmetric KCl solution, the two states had slope conductances of 163 6 9 pS and 80 6 8 pS (Fig. 5B, closed and open circles, respectively). With K-gluconate solution in the bath and KCl solution in the pipette, only inward currents were observed (Fig. 5B, triangles for the upper conductance level). Therefore, the channel was selective for Cl2. Agonists of IP3Rs (IP3, Ca21, and ATP at 2 mM, 0.26 mM, and 0.5 mM, respectively) activated ion channels in 7 out of
NUCLEAR ION CHANNELS OF HIPPOCAMPAL NEURONS
873
express ion channels with different biophysical properties, suggesting different functional specialization of the two membranes. The nuclear membranes of hippocampal neurons contained (1) a variety of spontaneously active cation and anion channels and (2) ligand-gated Ca21-release channels. The main properties of these channels are summarized in Table 1. Of all nuclear ion channels only two, IP3Rs and LCC channels were found in the nuclear membrane at high density. The density of the other nuclear channels was low. Molecular identity was unequivocally established only for the ligand-gated channels, IP3Rs and RyRs. As discussed below, the two-subconductance state Cl2 channel of DG granule neurons resembles two-pore ClC channels. The molecular identities of the other channels are unknown. The functional role of nuclear ion channels is not well understood. The nuclear envelope is a part of the ER and there is a growing body of evidence that it is a functional Ca21 store (Nicotera et al., 1990; Gerasimenko et al., 1995, 2003; Stehno-Bittel et al., 1995; Marchenko et al., 2003; Marchenko and Thomas, 2006). Indeed, numerous IP3Rs were found in the nuclear membranes of CA1 pyramidal and, to a lesser degree, DG granule neurons. In neurons with high levels of
FIGURE 4. Anion channels in the outer nuclear membrane of DG granule cells. (A) An original recording of the anion channel at holding potentials of 140 mV and 240 mV (the zero current levels are indicated by an arrow). (B) A current–voltage relationship of the channel in symmetric KCl solution (closed circles) and with K-Glu containing solution in the bath (open circles).
46 patches of the inner nuclear membrane (15%). In symmetric KCl solution the slope conductance of the channels was 378 6 11 pS (n 5 4). By their pattern of activity, conductance, and localization, they were similar to IP3Rs of CA1 pyramidal neurons. Agonists of RyRs (cADP ribose 10 mM or ryanodine 20 nM) were ineffective in 58 patches of the inner nuclear membrane, but ryanodine activated a channel in a single patch of the outer nuclear membrane of granule neurones (Fig. 6A). The slope conductance of the channel in symmetric KCl solution was 719 pS (Fig. 6B).
DISCUSSION The major finding of this work is that different types of hippocampal neuron express strikingly different sets of ion channels in their nuclear membranes. These data suggest that besides its function of separating genetic apparatus from the cytoplasm, the nuclear envelope carries out some additional functions, which may be different in different cells. We have also found that the inner and the outer nuclear membranes
FIGURE 5. Anion channels in the inner nuclear membrane of DG granule cells. (A) An original recording of the anion channel at holding potentials of 160 mV and 260 mV (an arrow indicates the close state of the channel and dash lines indicate different conductance levels of the two-level ion channel). (B) A current–voltage relationship of the lower (open circuses) and the higher (closed circles) conductance levels of the channel in symmetric KCl solution and with Glu-K solution in the bath (triangles). Hippocampus
874
FEDORENKO AND MARCHENKO
FIGURE 6. RyRs of the outer nuclear membrane of DG granule cells. (A) Recoding of RyRs at holding potential of 160 mV (an arrow indicates the close state of the channel). (B) A current– voltage relationship of the channel in symmetric KCl solution.
expression of IP3Rs they were mainly localized in the inner nuclear membrane, suggesting that the nuclear envelope of these neurons is specialized to release Ca21 into the nucleoplasm. In contrast, very few IP3Rs were detected in the nuclear membrane of CA3 pyramidal neurons and all of them were recorded in the outer nuclear membrane.
All hippocampal neurons express RyRs (Sharp et al., 1993, 1999; Hertle and Yeckel, 2007). Nevertheless, an RyR was recorded only in a single patch of the outer nuclear membrane of DG granule neurons. This observation indicates that of the two major types of intracellular Ca21-release channels only IP3Rs are targeted to the nuclear membrane in hippocampal neurons. The reason for this may be different modes of regulation of the receptors. IP3, required for activation of IP3Rs, can enter the nucleus from the cytoplasm, but it can also be synthesized in the nucleus (Irvine, 2000; Visnjic and Banfic, 2007). This allows regulation of nuclear IP3Rs independently. Besides, RyRs require for their activation much higher levels of Ca21 than IP3Rs (10 mM and 0.05 mM, respectively), which can be expected near the plasma membrane rather than in the depth of the cell. The main spontaneously active ion channels in hippocampal neurons were LCC channels. The density of these channels closely correlated with the density of nuclear IP3Rs—it was much higher in CA1 pyramidal and DG granule neurons than in CA3 pyramidal neurons and in neurons with high levels of nuclear IP3Rs they were mainly localized in the inner nuclear membrane. Therefore, LCC channels and IP3Rs are coexpressed and are apparently functionally connected. Ca21 release from the intracellular store as well as Ca21 uptake by Ca21ATPase leads to a massive transfer of positive charge from/into the lumen of the store and creates a charge imbalance that hampers further Ca21 movements. Spontaneously active cation channels may provide a route for counterflow of K1 to reduce changes in potential between the lumen of the nuclear envelope and the nucleoplasm, thereby facilitating Ca21 release and uptake (Marchenko et al., 2005; Marchenko and Thomas, 2006; Yamashita et al., 2006). It has been argued that K1 inflow via RyRs and probably IP3Rs is sufficient to compensate for Ca21 outflow during Ca21 release (Gillespie and Fill, 2008). Nevertheless, in all studied neurons, the density of LCC channels was significantly higher than IP3Rs. For example, in CA1 pyramidal neurons LCC channels were found in 37% of patches with the density
TABLE 1. Ion Channels in the Nuclear Membrane of Hippocampal Neurons Cell type CA1 pyramidal neurons CA3 pyramidal neurons DG granule neurons
a
Conductancea 248 6 6 366 6 5 156 6 4 210 6 6
pS pS pS pS
179 6 15 pS 378 6 11 pS 719 pS 305 6 16 pS 80 6 8 pS and 163 6 9 pS
A slope conductance in symmetric KCl solution.
Hippocampus
Selectivity
Localization
Identity
Density
R1 Ca21 Cl2 R1
Mainly the inner nuclear membrane Mainly the inner nuclear membrane The outer nuclear membrane Both the inner and the outer nuclear membranes
Unknown IP3Rs Unknown Unknown
High High Low Very Low
R1 Ca21 Ca21 Cl2 Cl2
Mainly the inner nuclear membrane Mainly the inner nuclear membrane The outer nuclear membrane The outer nuclear membrane Inner nuclear membrane
Unknown IP3Rs RyRs Unknown ClC?
High High Very Low Low Low
NUCLEAR ION CHANNELS OF HIPPOCAMPAL NEURONS of 3.8 channels per patch whereas only 25% of patches contained IP3Rs with much lower density (2.1 channels per patch). Taking into account that conductance of LCC channels was only 32% lower than IP3Rs, they may account for the most part of compensatory K1 counterflow during Ca21 release. Moreover, near 0 mV LCC channels stay about constantly open (Po 5 0.8) thereby preventing any changes in the lumen membrane potential due to Ca21 movements (Fedorenko et al., 2010). The prevailing localization of LCC channels in the inner nuclear membrane may prevent transient differences in potential between the nucleoplasm and the cytoplasm during Ca21 release from the nuclear envelope. As the nuclear envelope is a part of the ER, we can assume that LCC channels play the same function in the rest of the ER. LCC channels in different hippocampal neurons had high, but different conductances, but in other respects, they very much resembled each other— they had high Po, slow kinetics, voltage-dependence, and selectivity for monovalent cations. Channels with very similar properties were previously reported in cerebellar Purkinje neurons (Marchenko et al., 2005). Similarities in the properties of nuclear LCC channels suggest that they may be members of a family of structurally related intracellular channels of so far unknown molecular identity. The density of anion channels in the nuclear membranes of hippocampal neurons was much lower than the density of LCC channels, but they were much more variable. They differed in conductance, pattern of activity and voltagedependence (Table 1). One of the anion channels recorded in the nuclear membrane of DG neurons quickly fluctuated between two similar subconductance states, thereby resembling members of the ClC family of ion channels (Jentsch et al., 2001; Stauber et al., 2012). All members of the ClC family are dimers forming two-pore (‘double-barreled’) anion channels. Many members of the ClC family (ClC-3–ClC-7) are mostly located in the membranes of intracellular organelles. The molecular identities of other nuclear anion channels are unknown. A probable general function of anion channels of the nuclear envelope is to maintain anion balance between the lumen and cyto/nucleoplasm. As many members of the ClC family are Cl2/H1 exchangers (Stauber et al., 2012), nuclear ClC channels may be involved in pH regulation in the lumen of the nuclear envelope. Our results suggest that: (1) IP3Rs, but not RyRs are targeted to the nuclear membrane of particular types of hippocampal neurons; (2) the density of nuclear IP3Rs correlates with the general level of expression of the receptors in a given type of neuron; (3) in neurons with high levels of IP3Rs they are localized predominantly in the inner nuclear membrane; (4) the nuclear envelope of CA1 pyramidal and DG granule neurons seems to be specialized to release Ca21 into the nucleoplasm, which may amplify Ca21 signals entering the nucleus from the cytoplasm or generate nuclear Ca21 transients on its own; (5) the nuclear membranes of hippocampal neurons possess a set of spontaneously active cation and anion channels that is specific in each cell type; (6) LCC channels are
875
coexpressed with IP3Rs and are an integral part of the Ca21releasing machinery, providing a route for counterflow of R1 and thereby facilitating Ca21 movements.
REFERENCES Amaral DG, Lavenex P. 2007. Hippocampal neuroanatomy. In: Andersen P, Morris R, Amaral D, Bliss T, O’Keefe J, editors. The Hippocampus Book. New York: Oxford University Press. p 872. Bengtson CP, Bading H. 2012. Nuclear calcium signaling. Adv Exp Med Biol 970:377–405. Bezprozvanny I. 2005. The inositol 1,4,5-trisphosphate receptors. Cell Calcium 38(3–4):261–272. Dick O, Bading H. 2010. Synaptic activity and nuclear calcium signaling protect hippocampal neurons from death signal-associated nuclear translocation of FoxO3a induced by extrasynaptic N-methyl-D-aspartate receptors. J Biol Chem 285(25):19354–61. Fedorenko O, Yarotskyy V, Duzhyy D, Marchenko S. 2010. The large conductance ion channels in the nuclear envelope of central neurons. Pflugers Arch Eur J Physiol 460:1045–1050. Fedorenko OA, Marchenko SM. 2010. Spontaneously active ion channels of the nuclear envelope membranes. Fiziol Zh 56(5):95–105. Gerasimenko OV, Gerasimenko JV, Tepikin AV, Petersen OH. 1995. ATP-dependent accumulation and inositol trisphosphate- or cyclic ADP-ribose-mediated release of Ca21 from the nuclear envelope. Cell 80:439–444. Gerasimenko J, Maruyama Y, Tepikin A, Petersen OH, Gerasimenko O. 2003. Calcium signalling in and around the nuclear envelope. Biochem Soc Trans 31(Pt 1):76–78. Gillespie D, Fill M. 2008. Intracellular calcium release channels mediate their own countercurrent: the ryanodine receptor case study. Biophys J 95:3706–3714. Jentsch TJ, Stein V, Weinreich F, Zdebik AA. 2001. Molecular structure and physiological function of chloride channels. Physiol Rev 82:503–568. Hertle DN, Yeckel MF. 2007. Distribution of inositol-1,4,5trisphosphate receptor and ryanodine receptor isotypes during maturation of the rat hippocampus. Neuroscience 150:625–638. Humbert JP, Matter N, Artault JC, K€oppler P, Malviya AN. 1996. Inositol 1,4,5-trisphosphate receptor is located to the inner nuclear membrane vindicating regulation of nuclear calcium signaling by inositol 1,4,5-trisphosphate. J Biol Chem 271(1):478–485. Irvine R. 2000. Nuclear lipid signaling. Sci STKE 5(48):re1. Mak DO, McBride S, Foskett JK. 1998. Inositol 1,4,5-trisphosphate [correction of tris-phosphate] activation of inositol trisphosphate [correction of tris-phosphate] receptor Ca21 channel by ligand tuning of Ca21 inhibition. Proc Natl Acad Sci USA 95(26): 15821–15825. Marchenko SM, Yarotskyy VV, Kostyuk PG, Thomas RC. 2003. The nuclear envelope of cerebellar Purkinje cells is a calcium store with inositol 1,4,5-trisphosphate receptors restricted to the internal nuclear membrane. Ann SfN Meeting, Program No. 252.13. Marchenko SM, Yarotskyy VV, Kovalenko TN, Kostyuk PG, Thomas R. 2005. Spontaneously active and InsP3-activated ion channels in cell nuclei from rat cerebellar Purkinje and granule neurons. J Physiol 15(565):897–910. Marchenko SM, Thomas RC. 2006. Nuclear Ca21 signalling in cerebellar Purkinje neurons. Cerebellum 5(1):36–42. Mauceri D, Freitag HE, Oliveira AM, Bengtson CP, Bading H. 2011. Nuclear calcium-VEGFD signaling controls maintenance of dendrite arborization necessary for memory formation. Neuron 71(1): 117–130. Hippocampus
876
FEDORENKO AND MARCHENKO
Mazzanti M, Bustamante JO, Oberleithner H. 2001. Electrical dimension of the nuclear envelope. Physiol Rev 81(1):1–19. Miedema H. 2002. Surface potentials and the calculated selectivity of ion channels. Biophys J 82(1 Pt 1):156–159. Nicotera P, Orrenius S, Nilsson T, Berggren PO. 1990. An inositol 1,4,5-trisphosphate-sensitive Ca21 pool in liver nuclei. Proc Natl Acad Sci USA 87(17):6858–6862. Oliveira AM, Hemstedt TJ, Bading H. 2012. Rescue of agingassociated decline in Dnmt3a2 expression restores cognitive abilities. Nat Neurosci 15(8):1111–1113. Papadia S, Stevenson P, Hardingham NR, Bading H, Hardingham GE. 2005. Nuclear Ca21 and the cAMP response element-binding protein family mediate a late phase of activity-dependent neuroprotection. J Neuroscience 25(17):4279–4287. Rahman T, Skupin A, Falcke M, Taylor CW. 2009. Clustering of IP3 receptors by IP3 retunes their regulation by IP3 and Ca21. Nature 458(7238):655–659. Sharp AH, McPherson PS, Dawson TM, Aoki C, Campbell KP, Snyder SH. 1993. Differential immunohistochemical localization of inositol 1,4,5-trisphosphate- and ryanodine-sensitive Ca21 release channels in rat brain. J Neuroscience 13:3051–3063. Sharp AH, Nucifora FC, Blondel O, Sheppard CA, Zhang C, Snyder SH, Russell JT, Ryugo DK, Ross CA. 1999. Differential cellular
Hippocampus
expression of isoforms of inositol 1,4,5-triphosphate receptors in neurons and glia in brain. J Comp Neurol 406:207–220. Stauber T, Weinert S, Jentsch TJ. 2012. Cell biology and physiology of CLC chloride channels and transporters. Compr Physiol. 2(3): 1701–1744. Stehno-Bittel L, L€ uckhoff A, Clapham DE. 1995. Calcium release from the nucleus by InsP3 receptor channels. Neuron 14(1):163– 167. Visnjic D, Banfic H. 2007. Nuclear phospholipid signaling: phosphatidylinositol-specific phospholipase C and phosphoinositide 3-kinase. Pflugers Arch. 455(1):19–30. Yamashita M, Sugioka M, Ogawa Y. 2006. Voltage- and Ca21activated potassium channels in Ca21 store control Ca21 release. FEBS J 273(15):3585–3597. Zhang SJ, Zou M, Lu L, Lau D, Ditzel DA, Delucinge-Vivier C, Aso Y, Descombes P, Bading H. 2009. Nuclear calcium signaling controls expression of a large gene pool: identification of a gene program for acquired neuroprotection induced by synaptic activity. PLoS Genet 5(8):e1000604. Zhang SJ, Steijaert MN, Lau D, Sch€ utz G, Delucinge-Vivier C, Descombes P, Bading H. 2007. Decoding NMDA receptor signaling: identification of genomic programs specifying neuronal survival and death. Neuron 53(4):549–562.
Ion Channels of the Nuclear Membrane of Hippocampal Neurons Olena A. Fedorenko,1,2 and Sergey M. Marchenko1,2*
ABSTRACT: Rise in Ca21 concentration in the nucleus affects gene transcription and has been implicated in neuroprotection, transcriptiondependent neuronal plasticity, and pain modulation, but the mechanism of regulation of nuclear Ca21 remains poorly understood. The nuclear envelope is a part of the endoplasmic reticulum and may be one of the sources of nuclear Ca21. Here, we studied ion channels in the nuclear membrane of hippocampal neurons using the patch-clamp technique. We have found that the nuclear membrane of CA1 pyramidal and dentate gyrus granule (DG), but not CA3 pyramidal neurons, was enriched in functional inositol 1,4,5-trisphosphate receptors/Ca21-release channels (IP3Rs) localized mainly in the inner nuclear membrane. A single nuclear ryanodine receptor (RyR) has been detected only in DG granule neurons. Nuclei of the hippocampal neurons also expressed a variety of spontaneously active cation and anion channels specific for each type of neuron. In particular, large-conductance ion channels selective for monovalent cations (LCC) were coexpressed with IP3Rs. These data suggest that: (1) the nuclear membranes of hippocampal neurons contain distinct sets of ion channels, which are specific for each type of neuron; (2) IP3Rs, but not RyRs are targeted to the inner nuclear membrane of CA1 pyramidal and DG granule, but they were not found in the nuclear membranes of CA3 pyramidal neurons; (3) the nuclear envelope of these neurons is specialized to release Ca21 into the nucleoplasm which may amplify Ca21 signals entering the nucleus from the cytoplasm or generate Ca21 transients on its own; (4) LCC channels are an integral part the of Ca21-releasing machinery providing a route for counterflow of R1 and thereby facilitating Ca21 movement in and out of the Ca21 C 2014 Wiley Periodicals, Inc. store. V KEY WORDS: ion channels; cell nucleus; endoplasmic reticulum; calcium stores; hippocampus
INTRODUCTION The nuclear envelope is a specific organelle, which separates the genetic apparatus of a cell from the rest of the cytoplasm. It is a part of the endoplasmic reticulum (ER), connected to it by its outer nuclear
membrane and common luminal space. The relatively large size of the nuclear envelope enables direct patchclamp recording from the nuclear membranes. This approach has been successfully employed for investigation of inositol 1,4,5-trisphosphate receptors (IP3Rs) (Mak et al., 1998; Marchenko et al., 2003; Rahman et al., 2009). A variety of other ion channels was reported in the nuclear membrane (Mazzanti et al., 2001; Fedorenko et al., 2010; Fedorenko and Marchenko, 2010). Besides being a useful model of the ER, the nuclear envelope is involved in the regulation of nuclear Ca21. A rise in nuclear Ca21 concentration ([Ca21]n) affects the transcription of a large pool of genes in hippocampal neurons and has been implicated in various physiological phenomena including neuronal survival, aging and transcription-dependent long-lasting changes in synaptic efficiency (Papadia et al., 2005; Zhang et al., 2007, 2009; Dick and Bading, 2010; Mauceri et al., 2011; Bengtson and Bading, 2012; Oliveira et al., 2012). We have previously reported multiple functional IP3Rs and spontaneously active cation and anion channels in the nuclear membranes of cerebellar Purkinje and granule neurons (Marchenko et al., 2005). Expression of ion channels greatly varied in different types of cerebellar neurons and we, for example, have not detected IP3-activated channels in the nuclear membrane of cerebellar granule neurons. Therefore, every cell type has to be studied independently. Here, we studied ion channels in the membranes of nuclei isolated from CA1 and CA3 pyramidal and dentate gyrus (DG) granule neurons of the rat hippocampus.
MATERIALS AND METHODS 1
Bogomoletz Institute of Physiology, Department of Brain Physiology, 4 Bogomoletz Street, Kiev, 01024, Ukraine; 2 State Key Laboratory of Molecular and Cellular Biology, 4 Bogomoletz Street, Kiev, 01024, Ukraine Additional Supporting Information may be found in the online version of this article. Grant sponsor: The State Fund for Fundamental Researches; Grant number: DFFD F 46.2/001. *Correspondence to: Sergey M. Marchenko, Bogomoletz Institute of Physiology, Department of Brain Physiology, 4 Bogomoletz Street, Kiev 01024, Ukraine. E-mail: [email protected] Accepted for publication 31 March 2014. DOI 10.1002/hipo.22276 Published online 25 April 2014 in Wiley Online Library (wileyonlinelibrary.com). C 2014 WILEY PERIODICALS, INC. V
Isolation of Neuronal Nuclei All experimental procedures involving animals and their care were conducted in accordance with the European Communities Council Directive of 24 November 1986. Hippocampi were isolated from three-week-old male Wistar rats. Transverse slices of the hippocampus were cut by hand and each slice was placed in an eppendorf tube with 1 ml of ice-cold solution containing (mM): potassium gluconate 150, Hepes-KOH 10 (pH 7.2). Protease Inhibitor Cocktail (Roche Diagnostics Gmbh) was added to the solution
870
FEDORENKO AND MARCHENKO
according to the manufacturer’s instructions. The eppendorf tubes with hippocampal slices were placed in a freezer and used for experiments during the next 1–3 days. The pyramidal layers of the CA1 or CA3 areas or DG granule layer were carefully cut using microsurgical instruments. The resulting preparation was homogenized by passing through a 23 gauge needle. The homogenate was centrifuged, the pellet was resuspended in the same solution and placed into a working chamber on the stage of a Leica DMIRB inverted microscope (Leica Microsystems, Germany). The nuclei were allowed to attach to the glass bottom of the chamber and then debris was removed by washing with solution (hereafter referred to as KCl solution) containing (mM): KCl 150; HEPES 5, pH 7.2. Nuclei in the working chamber were visualized with the VX44 or Pixelfly CCD camera (PCO, Kelheim, Germany). To access the inner nuclear membrane the isolated nuclei were treated with 1% (w/v) sodium citrate (Humbert et al., 1996) and incubated for 15–20 min on ice while gently stirring (HI 200M magnetic stirrer, Hanna Instruments, 0.5–1 revolutions per second). They were then used in experiments as described above.
Electrophysiological Studies Single ion channels were recorded from nucleus-attached and excised patches of the nuclear membrane in the voltageclamp mode of the patch-clamp technique. There were no differences between data obtained in these configurations, so the data were pooled. Patch pipettes were prepared from borosilicate glass (Sutter Instrument, Novato, CA, USA) and filled, unless stated otherwise, with KCl solution. Pipette resistance ranged from 8 to 12 MX. Data acquisition was carried out using a Visual Patch VP-500 amplifier (Bio-Logic, Claix, France). Currents were filtered with a low-pass Bessel filter at 2 kHz, digitized at 10 kHz, and stored on a computer disk. All experiments were carried out at room temperature (18–20 C). In all recordings in this paper, the potential of the bath solution was considered to be 0 mV. Data are presented as the mean 6 standard error (s.e.m.) and n (number of observations). To determine the selectivities of ion channels, KCl solution was replaced with either K-gluconate solution (potassium gluconate 150 mM, HEPES 5, pH 7.2) or N-methyl-D-glucamine (NMDG) solution (NMDG-Cl 150 mM, HEPES 5, pH 7.2). Ca21 concentration was buffered with EGTA: EGTA 1 mM, CaCl2 0.6 mM; calculated free Ca21 concentration was 0.26 m. Solutions with agonists of IP3Rs or RyRs also contained ATP (0.5 mM).
RESULTS Isolated Nuclei of Hippocampal Neurons The heterogenous cell composition of the brain complicates isolation of nuclei from specific types of neurons. The regular anatomical structure of the hippocampus helps to overcome Hippocampus
FIGURE 1. Isolated nuclei of hippocampal neurons and anion channels in the outer nuclear membrane of CA1 pyramidal neurons. Cell nuclei isolated from CA1 (A) and CA3 (B) pyramidal neurons of the hippocampus and from granule cells of the dentate gyrus (C). CCD images of unstained nuclei in KCl solution, objective 3100. (D) Recording of the anion channels of the outer nuclear membrane of CA1 pyramidal neurons (an arrow indicates the close state of the channel). (E) A current–voltage relationship of anion channels in the symmetric 150 mM KCl solution (closed circles), and with low Cl– solution (130 mM K-Glu and 20 mM KCl) in the bath solution (open circles).
this problem. The CA1 and CA3 pyramidal layers of Ammon’s Horn and the granule layer of the Dentate Gyrus (DG) are composed of densely packed cell bodies of pyramidal and granule neurons respectively, with few other cell types (Amaral and Lavenex, 2007). In accordance with this, nuclei isolated from CA1, CA3 pyramidal, or DG layers were morphologically homogenous. Nuclei from CA1 pyramidal neurons were spherical in shape and 10–12 mm in diameter. They had one or more distinct dark nucleoli and clear light nucleoplasm (Fig. 1A). These morphological features and size set them apart from the nuclei of other cell types in the CA1 area. The nuclei isolated from CA3 pyramidal neurons were larger (15 lm), had spherical or ellipsoidal shape, several (2–4) clearly visible dark nucleoli and less light nucleoplasm (Fig. 1B). Nuclei from DG granule neurons had diameters of 7–8 mm, small nucleoli, and turbid nucleoplasm (Fig. 1C).
NUCLEAR ION CHANNELS OF HIPPOCAMPAL NEURONS
871
Patch-clamp recordings of the nuclear membranes from these three types of hippocampal neurons revealed multiple ion channels, that varied in their selectivity, conductance, and pattern of activity.
Ion Channels in the Nuclear Membranes of CA1 Pyramidal Neurons In the outer nuclear membrane of ff1 pyramidal neurons, we recorded spontaneously active ion channels with conductances of 156 6 4 pS (n 5 3) in symmetric Kl solution (Fig. 1E). The density of these channels was low (3 out of 56 patches). The channels demonstrated clear voltage-dependence (Fig. 1D). At positive potentials (and therefore negative potentials at the luminal side of the membrane) the open probability of these channels was low (0.02 at 140 mV), but significantly increased at negative potentials (0.08 at 240 mV). Replacement of 80% (130 mM) of KCl in the bath solution with potassium gluconate shifted the reversal potential of the current to 49.3 mV (Fig. 1E). The calculated reversal potential for Cl2 ions in these conditions was 50.9 mV. Replacement of KCl solution in the working chamber with NMDG-Cl solution did not affect the channel. Therefore, the channels were permeable for Cl2 and practically impermeable to K1. No such channels were detected in the inner nuclear membrane. The most common spontaneously active ion channels found in the both inner and outer nuclear membranes of CA1 pyramidal neurons were large-conductance cation (LCC) channels. These channels were mainly localized in the inner nuclear membrane. Out of 151 patches of the inner nuclear membrane, 56 contained one or more (up to 7, mean 3.8 channels per patch) of these channels (37% of patches) and very few were found in the outer nuclear membrane (1 out of 56, or less than 2%). The channels demonstrated clear voltagedependence. At positive potentials at the luminal side of the membrane, they were mainly in the open state and their open probability (Po) was 0.8. At negative potentials, the activity of the channels significantly declined (Fig. 2A). Long exposure to negative potentials below 240 mV could completely, but reversibly block the channels. In symmetric KCl solution, the slope conductance of the channel was 248 6 6 pS (n 5 8) (Fig. 2B). To determine ion selectivity of the channel, KCl solution in the working chamber was replaced with NMDG solution; the patch pipettes contained the standard KCl solution. Under these conditions, only outward currents were recorded over the range of potentials from 280 mV to 180 mV. Equimolar replacement of KCl in the outer solution with NaCl did not noticeably affect the channel. When patch pipettes were filled with 100 mM BaCl2 or CaCl2 solution, only inward currents were observed. These data suggest that LCC channels were selective to monovalent (K1 and Na1) cations and impermeable to divalent (Ca21, Ba21) cations. Agonists of IP3Rs (IP3 1–10 mM, Ca21 0.26 mM, and ATP 0.5 mM) applied to the nucleoplasmic (bath) side of the inner nuclear membrane of CA1 pyramidal neurons activated highconductance channels in 25% of patches (37 out of 151
FIGURE 2. Large conductance cation (LCC) channels in the inner nuclear membrane of CA1 pyramidal neurons. (A) LCC channel recorded at holding potentials of 160 mV and 260 mV. (B) A current–voltage relationship of LCC channels in symmetric KCl solution (closed circles), with NMDG solution in the bath (triangles), and with 100 mM CaCl2 in the patch-pipette (open circles). a21 either in pipette or bath solution at large concentrations (10 mM) reduced the channel amplitude.
patches, on the average 2.1 channels per patch). In the outer nuclear membrane of CA1 neurons, agonists of IP3Rs in the pipette solution (the cytoplasmic side of the outer nuclear membrane) activated channels of the same conductance only in a few patches (3 out of 54 patches, or 5.6%). These channels often overlapped with LCC channels in the same patch. IP3 activated the channels at concentrations 50 nM and saturated their activity (Po 95% of Po at 10 mM IP3) at concentrations of 2 mM (Fig. 3A). When added to the patch pipette solution (the luminal side of the inner nuclear membrane), the agonists did not activate any channels (n 5 37). So, the ligandbinding domains of these IP3Rs were directed toward the nucleoplasm. In symmetric KCl solution, IP3/Ca21-activated channels had linear current–voltage relationships with slope conductances of 366 6 5 pS (n 5 6, Fig. 3B, closed circles). When patch pipettes were filled with solution containing 50 mM CaCl2 and 30 mM KCl, the reversal potential of IP3-activated channels was 224.3 6 2.1 mV (n 5 4, Fig. 3B, open circles). This reversal potential corresponds to a relative permeability of the channel to Ca21 (PCa/PK) of 6.6 (Miedema, 2002). IP3-activated channels were also permeable to Ba21. When patch pipettes were filled with solution containing 100 mM BaCl2, the reversal potential of IP3-activated channels was Hippocampus
872
FEDORENKO AND MARCHENKO nuclear membrane. Large-conductance spontaneously active ion channels, resembling LCC channels of CA1 pyramidal neurons, were recorded in two patches of the outer membrane and two patches of the inner nuclear membrane. In symmetric KCl solution, the channel conductance was 210 6 6 pS (n 5 4). The channels were voltage-dependent, being activated by positive potentials and reversibly inhibited by negative potentials at the luminal side of the nuclear membrane. The rarity of these channels did not allow us to study their properties in more detail. To record Ca21 channels, the pipettes or bath were filled with solutions that contained agonists of RyRs (ryanodine 10 nM) or IP3Rs (IP3, 2 mM; Ca21, 0.26 mM). Two IP3Rs were recorded in a single patch (out of 75) of the outer nuclear membrane. No IP3R or RyRs were detected in the inner nuclear membrane (n 5 86).
Ion Channels in the Nuclear Membranes of DG Granule Neurons
FIGURE 3. IP3Rs in the inner nuclear membrane of CA1 pyramidal neurons. (A) IP3Rs were activated by IP3 in the presence of Ca21 (250 nM), the patch contained at least two IP3Rs, holding potential was 140 mV. (B) A current–voltage relationship of IP3Rs of CA1 pyramidal neurons in symmetric KCl solution was linear with a slope conductance of 366 6 5 pS (closed circles); when patch pipettes were filled with high Ca1-containing solution (50 mM CaCl2 and 30 mM KCl) the reversal potential of IP3/ Ca21-activated channels was 224.3 6 2.1 mV indicating Ca21 selectivity of the channel with PCa/PK 5 6.6 (open circles).
233.1 6 2.8 mV (n 5 4, not illustrated). The calculated PBa/ PK was 6.5. Therefore, IP3/Ca21-activated nuclear channels in CA1 pyramidal neurons can be identified as IP3Rs with properties similar to those previously reported for IP3Rs (Bezprozvanny, 2005; Marchenko et al., 2005). Ryanodine receptors/Ca21-release channels were reported to be expressed in CA1 pyramidal neurons (Sharp et al., 1993, 1999; Hertle and Yeckel, 2007). Nevertheless the agonists of RyRs, cADP ribose (10 mM) or ryanodine (10 nM), added to both pipette and bath solutions did not activate any channels in either the inner or the outer nuclear membrane of CA1 pyramidal neurons (n > 40 for each membrane).
Ion Channels in the Nuclear Membranes of CA3 Pyramidal Neurons The nuclear membranes of CA3 pyramidal neurons were characterized by extremely low densities of ion channels. We detected channel activity in only 3 out of 198 patches of the outer membrane and in only 2 out of 204 patches of the inner Hippocampus
Patch-clamp recording from the nuclear membranes of DG granule neurons revealed a variety of ion channels with densities lower than in the CA1 area, but higher than in CA3. A channel with high amplitude was recorded in the outer nuclear membrane of granule neurons (Fig. 4A). In symmetric KCl solution, it had a slope conductance of 305 6 16 pS (n 5 5, Fig. 4B, closed circles). Substitution of KCl solution with NMDG-Cl solution did not affect the channel, but with R1gluconate in the bath only inward currents were observed (Fig. 4B, open circles), suggesting that the channel was selective for Cl2. The most common spontaneously active channels in the nuclear membranes of DG granule neurons were largeconductance ion channels. These channels by their pattern of activity, high amplitude, slow kinetics, voltage-dependence, and localization mainly in the inner nuclear membrane resembled LCC channels of CA1 pyramidal neurons, but had smaller conductance. In symmetric KCl solution, the slope conductance of the channels was 179 6 15 pS (n 5 6). Equimolar substitution of KCl in the bath solution with K-gluconate did not affect the channels, but when KCl bath solution was replaced with NMDG-Cl solution only outward currents were observed over the range of potentials from 280 mV to 180 mV. With CaCl2 (100 mM) in the pipette only outward currents were detected. Therefore, the channel was permeable to K1, but impermeable to Ca21, Cl2, and NMDG. One more spontaneously active ion channel was recorded in the inner nuclear membrane of granule neurons. The channel quickly fluctuated between two-subconductance states (Fig. 5A). In symmetric KCl solution, the two states had slope conductances of 163 6 9 pS and 80 6 8 pS (Fig. 5B, closed and open circles, respectively). With K-gluconate solution in the bath and KCl solution in the pipette, only inward currents were observed (Fig. 5B, triangles for the upper conductance level). Therefore, the channel was selective for Cl2. Agonists of IP3Rs (IP3, Ca21, and ATP at 2 mM, 0.26 mM, and 0.5 mM, respectively) activated ion channels in 7 out of
NUCLEAR ION CHANNELS OF HIPPOCAMPAL NEURONS
873
express ion channels with different biophysical properties, suggesting different functional specialization of the two membranes. The nuclear membranes of hippocampal neurons contained (1) a variety of spontaneously active cation and anion channels and (2) ligand-gated Ca21-release channels. The main properties of these channels are summarized in Table 1. Of all nuclear ion channels only two, IP3Rs and LCC channels were found in the nuclear membrane at high density. The density of the other nuclear channels was low. Molecular identity was unequivocally established only for the ligand-gated channels, IP3Rs and RyRs. As discussed below, the two-subconductance state Cl2 channel of DG granule neurons resembles two-pore ClC channels. The molecular identities of the other channels are unknown. The functional role of nuclear ion channels is not well understood. The nuclear envelope is a part of the ER and there is a growing body of evidence that it is a functional Ca21 store (Nicotera et al., 1990; Gerasimenko et al., 1995, 2003; Stehno-Bittel et al., 1995; Marchenko et al., 2003; Marchenko and Thomas, 2006). Indeed, numerous IP3Rs were found in the nuclear membranes of CA1 pyramidal and, to a lesser degree, DG granule neurons. In neurons with high levels of
FIGURE 4. Anion channels in the outer nuclear membrane of DG granule cells. (A) An original recording of the anion channel at holding potentials of 140 mV and 240 mV (the zero current levels are indicated by an arrow). (B) A current–voltage relationship of the channel in symmetric KCl solution (closed circles) and with K-Glu containing solution in the bath (open circles).
46 patches of the inner nuclear membrane (15%). In symmetric KCl solution the slope conductance of the channels was 378 6 11 pS (n 5 4). By their pattern of activity, conductance, and localization, they were similar to IP3Rs of CA1 pyramidal neurons. Agonists of RyRs (cADP ribose 10 mM or ryanodine 20 nM) were ineffective in 58 patches of the inner nuclear membrane, but ryanodine activated a channel in a single patch of the outer nuclear membrane of granule neurones (Fig. 6A). The slope conductance of the channel in symmetric KCl solution was 719 pS (Fig. 6B).
DISCUSSION The major finding of this work is that different types of hippocampal neuron express strikingly different sets of ion channels in their nuclear membranes. These data suggest that besides its function of separating genetic apparatus from the cytoplasm, the nuclear envelope carries out some additional functions, which may be different in different cells. We have also found that the inner and the outer nuclear membranes
FIGURE 5. Anion channels in the inner nuclear membrane of DG granule cells. (A) An original recording of the anion channel at holding potentials of 160 mV and 260 mV (an arrow indicates the close state of the channel and dash lines indicate different conductance levels of the two-level ion channel). (B) A current–voltage relationship of the lower (open circuses) and the higher (closed circles) conductance levels of the channel in symmetric KCl solution and with Glu-K solution in the bath (triangles). Hippocampus
874
FEDORENKO AND MARCHENKO
FIGURE 6. RyRs of the outer nuclear membrane of DG granule cells. (A) Recoding of RyRs at holding potential of 160 mV (an arrow indicates the close state of the channel). (B) A current– voltage relationship of the channel in symmetric KCl solution.
expression of IP3Rs they were mainly localized in the inner nuclear membrane, suggesting that the nuclear envelope of these neurons is specialized to release Ca21 into the nucleoplasm. In contrast, very few IP3Rs were detected in the nuclear membrane of CA3 pyramidal neurons and all of them were recorded in the outer nuclear membrane.
All hippocampal neurons express RyRs (Sharp et al., 1993, 1999; Hertle and Yeckel, 2007). Nevertheless, an RyR was recorded only in a single patch of the outer nuclear membrane of DG granule neurons. This observation indicates that of the two major types of intracellular Ca21-release channels only IP3Rs are targeted to the nuclear membrane in hippocampal neurons. The reason for this may be different modes of regulation of the receptors. IP3, required for activation of IP3Rs, can enter the nucleus from the cytoplasm, but it can also be synthesized in the nucleus (Irvine, 2000; Visnjic and Banfic, 2007). This allows regulation of nuclear IP3Rs independently. Besides, RyRs require for their activation much higher levels of Ca21 than IP3Rs (10 mM and 0.05 mM, respectively), which can be expected near the plasma membrane rather than in the depth of the cell. The main spontaneously active ion channels in hippocampal neurons were LCC channels. The density of these channels closely correlated with the density of nuclear IP3Rs—it was much higher in CA1 pyramidal and DG granule neurons than in CA3 pyramidal neurons and in neurons with high levels of nuclear IP3Rs they were mainly localized in the inner nuclear membrane. Therefore, LCC channels and IP3Rs are coexpressed and are apparently functionally connected. Ca21 release from the intracellular store as well as Ca21 uptake by Ca21ATPase leads to a massive transfer of positive charge from/into the lumen of the store and creates a charge imbalance that hampers further Ca21 movements. Spontaneously active cation channels may provide a route for counterflow of K1 to reduce changes in potential between the lumen of the nuclear envelope and the nucleoplasm, thereby facilitating Ca21 release and uptake (Marchenko et al., 2005; Marchenko and Thomas, 2006; Yamashita et al., 2006). It has been argued that K1 inflow via RyRs and probably IP3Rs is sufficient to compensate for Ca21 outflow during Ca21 release (Gillespie and Fill, 2008). Nevertheless, in all studied neurons, the density of LCC channels was significantly higher than IP3Rs. For example, in CA1 pyramidal neurons LCC channels were found in 37% of patches with the density
TABLE 1. Ion Channels in the Nuclear Membrane of Hippocampal Neurons Cell type CA1 pyramidal neurons CA3 pyramidal neurons DG granule neurons
a
Conductancea 248 6 6 366 6 5 156 6 4 210 6 6
pS pS pS pS
179 6 15 pS 378 6 11 pS 719 pS 305 6 16 pS 80 6 8 pS and 163 6 9 pS
A slope conductance in symmetric KCl solution.
Hippocampus
Selectivity
Localization
Identity
Density
R1 Ca21 Cl2 R1
Mainly the inner nuclear membrane Mainly the inner nuclear membrane The outer nuclear membrane Both the inner and the outer nuclear membranes
Unknown IP3Rs Unknown Unknown
High High Low Very Low
R1 Ca21 Ca21 Cl2 Cl2
Mainly the inner nuclear membrane Mainly the inner nuclear membrane The outer nuclear membrane The outer nuclear membrane Inner nuclear membrane
Unknown IP3Rs RyRs Unknown ClC?
High High Very Low Low Low
NUCLEAR ION CHANNELS OF HIPPOCAMPAL NEURONS of 3.8 channels per patch whereas only 25% of patches contained IP3Rs with much lower density (2.1 channels per patch). Taking into account that conductance of LCC channels was only 32% lower than IP3Rs, they may account for the most part of compensatory K1 counterflow during Ca21 release. Moreover, near 0 mV LCC channels stay about constantly open (Po 5 0.8) thereby preventing any changes in the lumen membrane potential due to Ca21 movements (Fedorenko et al., 2010). The prevailing localization of LCC channels in the inner nuclear membrane may prevent transient differences in potential between the nucleoplasm and the cytoplasm during Ca21 release from the nuclear envelope. As the nuclear envelope is a part of the ER, we can assume that LCC channels play the same function in the rest of the ER. LCC channels in different hippocampal neurons had high, but different conductances, but in other respects, they very much resembled each other— they had high Po, slow kinetics, voltage-dependence, and selectivity for monovalent cations. Channels with very similar properties were previously reported in cerebellar Purkinje neurons (Marchenko et al., 2005). Similarities in the properties of nuclear LCC channels suggest that they may be members of a family of structurally related intracellular channels of so far unknown molecular identity. The density of anion channels in the nuclear membranes of hippocampal neurons was much lower than the density of LCC channels, but they were much more variable. They differed in conductance, pattern of activity and voltagedependence (Table 1). One of the anion channels recorded in the nuclear membrane of DG neurons quickly fluctuated between two similar subconductance states, thereby resembling members of the ClC family of ion channels (Jentsch et al., 2001; Stauber et al., 2012). All members of the ClC family are dimers forming two-pore (‘double-barreled’) anion channels. Many members of the ClC family (ClC-3–ClC-7) are mostly located in the membranes of intracellular organelles. The molecular identities of other nuclear anion channels are unknown. A probable general function of anion channels of the nuclear envelope is to maintain anion balance between the lumen and cyto/nucleoplasm. As many members of the ClC family are Cl2/H1 exchangers (Stauber et al., 2012), nuclear ClC channels may be involved in pH regulation in the lumen of the nuclear envelope. Our results suggest that: (1) IP3Rs, but not RyRs are targeted to the nuclear membrane of particular types of hippocampal neurons; (2) the density of nuclear IP3Rs correlates with the general level of expression of the receptors in a given type of neuron; (3) in neurons with high levels of IP3Rs they are localized predominantly in the inner nuclear membrane; (4) the nuclear envelope of CA1 pyramidal and DG granule neurons seems to be specialized to release Ca21 into the nucleoplasm, which may amplify Ca21 signals entering the nucleus from the cytoplasm or generate nuclear Ca21 transients on its own; (5) the nuclear membranes of hippocampal neurons possess a set of spontaneously active cation and anion channels that is specific in each cell type; (6) LCC channels are
875
coexpressed with IP3Rs and are an integral part of the Ca21releasing machinery, providing a route for counterflow of R1 and thereby facilitating Ca21 movements.
REFERENCES Amaral DG, Lavenex P. 2007. Hippocampal neuroanatomy. In: Andersen P, Morris R, Amaral D, Bliss T, O’Keefe J, editors. The Hippocampus Book. New York: Oxford University Press. p 872. Bengtson CP, Bading H. 2012. Nuclear calcium signaling. Adv Exp Med Biol 970:377–405. Bezprozvanny I. 2005. The inositol 1,4,5-trisphosphate receptors. Cell Calcium 38(3–4):261–272. Dick O, Bading H. 2010. Synaptic activity and nuclear calcium signaling protect hippocampal neurons from death signal-associated nuclear translocation of FoxO3a induced by extrasynaptic N-methyl-D-aspartate receptors. J Biol Chem 285(25):19354–61. Fedorenko O, Yarotskyy V, Duzhyy D, Marchenko S. 2010. The large conductance ion channels in the nuclear envelope of central neurons. Pflugers Arch Eur J Physiol 460:1045–1050. Fedorenko OA, Marchenko SM. 2010. Spontaneously active ion channels of the nuclear envelope membranes. Fiziol Zh 56(5):95–105. Gerasimenko OV, Gerasimenko JV, Tepikin AV, Petersen OH. 1995. ATP-dependent accumulation and inositol trisphosphate- or cyclic ADP-ribose-mediated release of Ca21 from the nuclear envelope. Cell 80:439–444. Gerasimenko J, Maruyama Y, Tepikin A, Petersen OH, Gerasimenko O. 2003. Calcium signalling in and around the nuclear envelope. Biochem Soc Trans 31(Pt 1):76–78. Gillespie D, Fill M. 2008. Intracellular calcium release channels mediate their own countercurrent: the ryanodine receptor case study. Biophys J 95:3706–3714. Jentsch TJ, Stein V, Weinreich F, Zdebik AA. 2001. Molecular structure and physiological function of chloride channels. Physiol Rev 82:503–568. Hertle DN, Yeckel MF. 2007. Distribution of inositol-1,4,5trisphosphate receptor and ryanodine receptor isotypes during maturation of the rat hippocampus. Neuroscience 150:625–638. Humbert JP, Matter N, Artault JC, K€oppler P, Malviya AN. 1996. Inositol 1,4,5-trisphosphate receptor is located to the inner nuclear membrane vindicating regulation of nuclear calcium signaling by inositol 1,4,5-trisphosphate. J Biol Chem 271(1):478–485. Irvine R. 2000. Nuclear lipid signaling. Sci STKE 5(48):re1. Mak DO, McBride S, Foskett JK. 1998. Inositol 1,4,5-trisphosphate [correction of tris-phosphate] activation of inositol trisphosphate [correction of tris-phosphate] receptor Ca21 channel by ligand tuning of Ca21 inhibition. Proc Natl Acad Sci USA 95(26): 15821–15825. Marchenko SM, Yarotskyy VV, Kostyuk PG, Thomas RC. 2003. The nuclear envelope of cerebellar Purkinje cells is a calcium store with inositol 1,4,5-trisphosphate receptors restricted to the internal nuclear membrane. Ann SfN Meeting, Program No. 252.13. Marchenko SM, Yarotskyy VV, Kovalenko TN, Kostyuk PG, Thomas R. 2005. Spontaneously active and InsP3-activated ion channels in cell nuclei from rat cerebellar Purkinje and granule neurons. J Physiol 15(565):897–910. Marchenko SM, Thomas RC. 2006. Nuclear Ca21 signalling in cerebellar Purkinje neurons. Cerebellum 5(1):36–42. Mauceri D, Freitag HE, Oliveira AM, Bengtson CP, Bading H. 2011. Nuclear calcium-VEGFD signaling controls maintenance of dendrite arborization necessary for memory formation. Neuron 71(1): 117–130. Hippocampus
876
FEDORENKO AND MARCHENKO
Mazzanti M, Bustamante JO, Oberleithner H. 2001. Electrical dimension of the nuclear envelope. Physiol Rev 81(1):1–19. Miedema H. 2002. Surface potentials and the calculated selectivity of ion channels. Biophys J 82(1 Pt 1):156–159. Nicotera P, Orrenius S, Nilsson T, Berggren PO. 1990. An inositol 1,4,5-trisphosphate-sensitive Ca21 pool in liver nuclei. Proc Natl Acad Sci USA 87(17):6858–6862. Oliveira AM, Hemstedt TJ, Bading H. 2012. Rescue of agingassociated decline in Dnmt3a2 expression restores cognitive abilities. Nat Neurosci 15(8):1111–1113. Papadia S, Stevenson P, Hardingham NR, Bading H, Hardingham GE. 2005. Nuclear Ca21 and the cAMP response element-binding protein family mediate a late phase of activity-dependent neuroprotection. J Neuroscience 25(17):4279–4287. Rahman T, Skupin A, Falcke M, Taylor CW. 2009. Clustering of IP3 receptors by IP3 retunes their regulation by IP3 and Ca21. Nature 458(7238):655–659. Sharp AH, McPherson PS, Dawson TM, Aoki C, Campbell KP, Snyder SH. 1993. Differential immunohistochemical localization of inositol 1,4,5-trisphosphate- and ryanodine-sensitive Ca21 release channels in rat brain. J Neuroscience 13:3051–3063. Sharp AH, Nucifora FC, Blondel O, Sheppard CA, Zhang C, Snyder SH, Russell JT, Ryugo DK, Ross CA. 1999. Differential cellular
Hippocampus
expression of isoforms of inositol 1,4,5-triphosphate receptors in neurons and glia in brain. J Comp Neurol 406:207–220. Stauber T, Weinert S, Jentsch TJ. 2012. Cell biology and physiology of CLC chloride channels and transporters. Compr Physiol. 2(3): 1701–1744. Stehno-Bittel L, L€ uckhoff A, Clapham DE. 1995. Calcium release from the nucleus by InsP3 receptor channels. Neuron 14(1):163– 167. Visnjic D, Banfic H. 2007. Nuclear phospholipid signaling: phosphatidylinositol-specific phospholipase C and phosphoinositide 3-kinase. Pflugers Arch. 455(1):19–30. Yamashita M, Sugioka M, Ogawa Y. 2006. Voltage- and Ca21activated potassium channels in Ca21 store control Ca21 release. FEBS J 273(15):3585–3597. Zhang SJ, Zou M, Lu L, Lau D, Ditzel DA, Delucinge-Vivier C, Aso Y, Descombes P, Bading H. 2009. Nuclear calcium signaling controls expression of a large gene pool: identification of a gene program for acquired neuroprotection induced by synaptic activity. PLoS Genet 5(8):e1000604. Zhang SJ, Steijaert MN, Lau D, Sch€ utz G, Delucinge-Vivier C, Descombes P, Bading H. 2007. Decoding NMDA receptor signaling: identification of genomic programs specifying neuronal survival and death. Neuron 53(4):549–562.
